JupiteX Get the app
Defence & Security18 Nov 2026 · about 7 min

Inside Modern Phishing Kits Built to Steal Microsoft 365 Sessions

The brief

A phishing kit is a ready-made set of code, templates, and instructions for running a phishing campaign and using stolen data. Basic kits may create a single fake login page. More advanced kits can include dashboards to track victims, change page content, and relay credentials in real time. This packaging lowers the effort needed to launch an attack. Some modern kits target Microsoft 365 session tokens rather than only passwords. For example, Kali365 can capture tokens through device-code flows or reverse-proxy lures, then replay them into live Microsoft 365 sessions. A stolen token can let an attacker enter a session where login and multifactor authentication are already complete. Huntress analyzed EvilTokens, Knight Office, and Kali365. Their capabilities range from capturing tokens to replaying sessions and maintaining access. That means defenders need to investigate what happened after a lure, including authentication, token use, device changes, and mailbox activity. Containment may require action on identity artifacts, not just removal of the phishing page or message.

01

What is a phishing kit, and how does it help an attacker steal Microsoft 365 access?

A phishing kit is a ready-made set of code, templates, and instructions for running a phishing campaign and using stolen data. Basic kits may create a single fake login page. More advanced kits can include dashboards to track victims, change page content, and relay credentials in real time. This packaging lowers the effort needed to launch an attack.

Some modern kits target Microsoft 365 session tokens rather than only passwords. For example, Kali365 can capture tokens through device-code flows or reverse-proxy lures, then replay them into live Microsoft 365 sessions. A stolen token can let an attacker enter a session where login and multifactor authentication are already complete.

Huntress analyzed EvilTokens, Knight Office, and Kali365. Their capabilities range from capturing tokens to replaying sessions and maintaining access. That means defenders need to investigate what happened after a lure, including authentication, token use, device changes, and mailbox activity. Containment may require action on identity artifacts, not just removal of the phishing page or message.

02

How many phishing kits did Huntress analyze, and what were they called?

Huntress researchers analyzed three modern phishing kits that target Microsoft 365 tokens. The kits are EvilTokens, Knight Office, and Kali365. This matters because stolen tokens can give attackers access to an authenticated session, rather than merely handing them a username and password to try.

EvilTokens appeared in a device-code phishing campaign that used Railway, a Platform-as-a-Service platform, as a token-harvesting and replay engine. Knight Office is an adversary-in-the-middle kit that steals valid Microsoft 365 session tokens and feeds them into an operator console. Kali365 combines token capture, replay, access maintenance, and other controls in one panel.

These examples show that phishing kits can offer different routes to the same dangerous result: usable sessions and potentially persistent access. The article describes device-code flows and reverse-proxy lures among Kali365’s methods, while Knight Office can help attackers set up persistence across tenant accounts. Defenders therefore need to investigate token use and account activity, not only the original phishing lure.

03

What is a session token, and why can it be more valuable to an attacker than a password?

A session token is the sign-in information that allows an authenticated session to continue. In the article’s examples, attackers capture authorization or sign-in tokens and use them to access an active session. The source focuses on token theft and replay; it does not provide a technical definition beyond those uses.

The key difference is what an attacker can do with the stolen item. A password may be only one part of signing in, while a replayed token can place an attacker in a session where the victim’s login and multifactor authentication are already complete. The article says attackers can then access cloud resources and the victim’s email.

This is why defenders should not treat a completed MFA prompt as proof that a session is trustworthy. The article describes kits that replay tokens into live Microsoft 365 sessions and maintain access. When token replay is suspected, investigators should follow how the token was used and review sign-ins, device changes, and mailbox activity.

04

How can device-code and adversary-in-the-middle attacks capture a token even when someone completes MFA?

Device-code phishing and adversary-in-the-middle attacks can steal access even when a victim completes MFA. The important point is that MFA may be completed in a flow whose resulting authorization or session information is captured by an attacker. A successful prompt alone does not prove that the resulting session is trustworthy.

In a device-code attack, the victim authenticates directly with the identity provider, while the attacker quietly captures and uses the resulting authorization. The victim may never type credentials into an unfamiliar page, making the experience feel like a normal sign-in. In an adversary-in-the-middle attack, the attacker relays the authentication flow through a fake page or intercepts an exchange in the legitimate app, then captures session information after the victim completes MFA.

The article says modern kits use these techniques to obtain usable authenticated sessions. Device-code authentication and token activity are therefore important investigation points. Defenders should follow the sequence from lure to authentication and token use, then check for new devices, authentication methods, and mailbox activity.

05

What can an attacker access by replaying a stolen Microsoft 365 session token?

A stolen Microsoft 365 session token can give an attacker more than the ability to attempt another login. Replaying it can place the attacker in an active session where the victim’s login and multifactor authentication are already complete. That makes the token a route to resources the victim can access.

The article states that attackers who replay a stolen token can gain access to the victim’s cloud resources and email accounts. It describes Kali365 replaying tokens into live Microsoft 365 sessions. Knight Office can also help attackers set up persistence across tenant accounts, while Kali365 includes features to maintain access.

The risk is therefore not limited to the moment a victim enters credentials on a phishing page. A usable session may continue to expose cloud resources and mailbox activity. Defenders are advised to investigate how a token was used, review sign-ins and mailbox activity, and address identity artifacts during containment. They should not assume that removing the original lure alone ends the compromise.

06

What signs can help defenders spot token theft, and what should they revoke or remove?

Token theft can leave clues across a sequence of events, so the article advises defenders to look beyond the phishing page. Signs include successful sign-ins from unfamiliar cloud infrastructure, device-code authentication, automated token refresh, new devices or authentication methods, and unusual mailbox activity. No single signal tells the whole story.

For example, an investigation can connect an unfamiliar successful sign-in with device-code authentication and later mailbox activity. The article recommends following the path from the lure through authentication, token issuance or theft, device registration, and mailbox activity. This timeline can help reveal identity compromise in the environment.

Containment should address identity artifacts, not only the original lure. The stated priorities are to revoke refresh tokens, remove unauthorized devices and Windows Hello credentials, and restrict or block device-code authentication where possible. These steps address ways an attacker may retain access after the phishing message or page is gone. The article does not provide a specific detection threshold or prescribe a single response sequence beyond these priorities.

07

Why do online services use session tokens to keep people signed in instead of asking for their password with every request?

Online services use session tokens to maintain a signed-in session between requests. As general background, the token can stand in for repeating the full sign-in process each time a person asks the service to do something. This makes continued use more convenient. The source article does not explain why services use tokens or describe their underlying design.

The article does explain the security consequence: attackers who obtain a sign-in token may replay it into an active session. In the Microsoft 365 examples, the login and MFA can already be complete, giving the attacker access to cloud resources and email. Device-code and adversary-in-the-middle techniques can capture authorization or session information during authentication.

Because the article focuses on defending against stolen tokens, it advises investigating sign-ins, token refresh, devices, authentication methods, and mailbox activity. It also recommends revoking refresh tokens and removing unauthorized devices and Windows Hello credentials. The broader implication is that protecting passwords alone may not address a compromised session; this final point is general background consistent with the article’s focus on token theft.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web